...others), define CEM by small upward rounding of reported net income that generates more than expected zeros and less than expected nines as second digit of earnings numbers. We partition our sample into positive and negative earnings to see whether the anomalous frequencies are reversed in negative earnings (losses) relative to positive earnings (profits), as predicted by CEM. In addition, we analyze net sales as a control variable. We find that, due to more ample opportunities to accounting manipulation, upward rounding is much more significant on the bottom line than on the top of an income statement. Consistent with prior studies, we find a reversed pattern of CEM for net losses. In addition, we find evidence for our expectation that CEM covaries with some institutional factors. We report that CEM decreases with spending on auditing, whereas it increases with the latitude of country's GAAP, its cultural values (power distance), and the importance of management bonus schemes. Contrary to our expectation, we do not find significant relationships between CEM and some factors commonly considered in related recent studies, such as the degree of shareholder protection or the alignment of financial and tax accounting.

Keywords: financial accounting; auditing; cosmetic earnings management; institutional factors; Benford's law; national culture.

JEL Classification: G15, G30, G38, M41.

I. INTRODUCTION

During the past few years, empirical research on earnings management behavior of firms has grown into one of the main streams in the field of financial accounting. Statistics from the ARN (1) working paper and accepted paper announcements indicate that while keywords earnings management had a hit rate of 11.5 percent among the 130 announcements published in 1996, the corresponding percentage was 19.4 percent for 2001 (221 announcements). Some of the most recent studies in the area have broken national boundaries by measuring earnings management around the world so as to find institutional settings that are associated with such behavior (Bhattacharya et al. 2003; Leuz et al. 2002). Furthermore, the research stream has matured to literature reviews and commentaries on the area (see e.g., Healey and Wahlen 1999; Parfet 2000; Beneish 2001; Stolowy and Breton 2000).

In this paper, we focus on cosmetic earnings management (hereafter CEM) around the world. By earnings management being cosmetic we mean a firm's tendency to do small upward rounding of reported net income, when such rounding yields an earnings number that seems abnormally larger than would be the case otherwise. For example, if the net income of a finn for a given year without rounding is $2.96 billion, then the firm exercises CEM by using discretionary accounting choices so that a net income number of, say, $3.01 billion is reported. The increase in reported earnings is only $0.05 billion (or 1.7 percent), but the resulting earnings number is perceived to be significantly larger because it is now in the "class" of $3 billion instead of $2 billion. (2) If the net income were smaller, say $2.68 billion, then the incentive to exercise earnings management of similar magnitude would be much smaller because it would yield an earnings number of only $2.73 billion, which remains in the same "class" of $2 billion.

Thus, CEM is expected to take place when the second digit of the unmanaged earnings number is nine. Then a small increase in earnings will suffice to increase the first digit of the earnings number by one, thus creating the intended cosmetic effect. Consequently, we can expect CEM to generate earnings numbers where there are a "deficit" of nines and a "surplus" of zeros as second digits.

Previous empirical studies have shown that firms in some countries do engage in this kind of CEM. The phenomenon was first detected by Carslaw (1988) in a sample of New Zealand firms. He reported a higher frequency of zeros and a lower frequency of nines as the second digit of reported earnings than could be expected under randomness. Subsequently, Thomas (1989) documented the same phenomenon from a large sample of U.S. firms. He also found that the frequency of second digits was reversed for negative net income numbers; when firms reported net losses, there were more nines and less zeros than could be expected. In addition, Niskanen and Keloharju (2000) analyzed earnings numbers of Finnish firms operating in a tax-driven financial accounting setting. Contrary to expectation, they found cosmetic earnings management to take place even in conditions where taxable income was based on reported earnings. This is somewhat surprising given that the alignment of financial and tax accounting makes upward rounding of earnings costly. More recently, Van Caneghem (2002) found evidence of CEM in U.K. firms. He documents that companies tend to round up reported earnings by increasing the first digit by one when they are faced with a nine as the second left-most digit. Consistent with expectation, such rounding could not be found for earnings before discretionary accruals. (3)

Carslaw (1988, 321-322) attributes the anomalies in the frequency of second digits to a psychological theory on cognitive reference points (see also, Van Caneghem [2002, 168], and the references cited therein). It suggests that human beings (users of financial statements) use yardsticks that are factors of ten in their assessment of numbers. Accordingly, a person tends to round an observed number to the nearest reference point in judging its magnitude. However, as the memory of human beings is limited and the process of rounding up is more complex than rounding down, people tend to store the most relevant information by assigning the largest weight to the first digit, then to the second, and so on, decaying through the other digits of the number. Thus, in human information processing and storing an earnings number of $2.96 billion is rounded down to $2.9 billion or $2 billion rather than up to its mathematical reference point of $3 billion.

Being aware of this psychological behavior of financial statement users, some firms may utilize it by managing earnings upward when the second digit of unmanaged earnings is nine. Thomas (1989, 773-774) argues that such small upward rounding can have a disproportionately large effect on the perceived profitability and value of the firm. This may provide managers with an incentive to exercise CEM. A small amount of earnings cosmetics may thus allegedly have a significant indirect economic impact.

Thomas (1989, 774) notes that CEM can also be motivated by contractual considerations parallel to the bonus scheme and debt covenant hypothesis of the positive accounting theory (Watts and Zimmerman 1986). When managers' bonus schemes, firms' debt covenants or internal budgets are defined in terms of round earnings numbers, large cash flow effects may be triggered by small-scale CEM. The explicit contractual linkages between round earnings numbers and firms' cash flows can thus provide a direct economic incentive for CEM.

Using a sample of 86,944 earnings observations from 21,662 firms in 18 countries for the five-year period 1995-1999, we document that, with a few exceptions, CEM tends to be a worldwide phenomenon. Following prior studies in the area, we measure CEM by the anomalous frequency of zeros and nines as second digits of earnings numbers. In addition, we partition our sample into positive and negative earnings to see whether the anomalous frequencies are reversed in negative numbers (losses) relative to positive numbers (profits), as suggested by Thomas (1989).

Unlike most prior studies, we use net sales as a control variable. We expect that cosmetic management of net sales numbers is less pronounced and less significant than management of earnings numbers. This is simply because the latitude in reporting net sales on the top line of an income statement through accounting method choices is presumably more restricted than the latitude in reporting net income on the bottom line. Apart from account receivables that relate to sales revenues, most accounting choices concerning discretionary current and noncurrent accruals (such as the valuation of inventories, or the accounting for goodwill amortization) have an impact only below the top line of the income statement. (4)

We hypothesize that the degree of CEM in individual countries is related to their institutional settings. Following Ali and Hwang (2000) and Leuz et al. (2002) among others, we examine shareholder protection, the alignment of financial and tax accounting, and the spending on auditing services as potential country-specific factors. In addition, we consider the idiosyncrasy of GAAP (using the International Accounting Standards as a benchmark), the importance of management bonus schemes, relative firm size, the value relevance of earnings, as well as the economic growth of our sample countries. Unlike any other related study, we also look at variables measuring certain dimensions of cultural values.

Our overall findings suggest that Miss Worm in Cosmetic Earnings Management is Spain, and the runners-up are Hong Kong and Singapore. While the order of other runners-up varies somewhat depending on the specific measure of CEM used, the general picture is very clear; the firms in these three countries clearly dominate the contest. At the other extreme of the continuum are Sweden, the U.K., and Norway; in these countries, among some others, CEM proves to be small and insignificant.

The results do not lend support to our expectation that the degree of CEM is associated with legal environment (shareholder protection) or the alignment of financial and tax accounting. We find no evidence to suggest that CEM is associated with either the value relevance of earnings or economic growth of our sample countries. Instead, we do find some support for the notion that CEM covaries with the latitude in country's GAAP, spending on auditing, the importance of management bonus schemes, and with cultural values (power distance). In an additional test, we find that our CEM variables are correlated with more traditional measures of earnings management, for example income smoothing.

In the following section of the paper, we describe the institutional factors we expect to be associated with the cross-country variation in CEM. The data and methods used in our empirical tests are explained in Section III, followed by our report of the empirical results in Section IV. Here we first report the aggregate results from each country, and then we document the findings from firm-specific tests and report the results from the tests concerning the associations of CEM with hypothesized institutional factors. We conclude with a brief summary of our main findings and their implications in Section V.

II. INSTITUTIONAL FACTORS

We expect several institutional factors to be associated with the variation of cosmetic earnings management across countries. The factors we consider relate to shareholder protection, properties of GAAP, taxation, the role of auditing, importance of management bonus schemes, finn size, as well as certain dimensions of national culture. In addition, we consider potential consequences of CEM such as the value relevance of earnings and the economic growth of a country. We now elaborate our expectations on these factors.

Shareholder Protection

Recent studies document that the jurisdictional arrangements and corporate governance systems in a country have an important role to play as a determinant of corporate finance decisions and the properties of financial accounting information. For example, LaPorta et al. (1997) suggest that the origin of the legal system and the degree of shareholder protection are significant explanatory variables for external market capitalization of firms' equity and for the intensity of initial public offerings. LaPorta et al. (2000) also document that these factors relate to dividend policies followed in different countries. In addition, the findings provided by Ball et al. (2000) and Ali and Hwang (2000) suggest that institutional factors relating to the legal environment (common versus code law system), sources of GAAP (governmental versus private sector standard-setting bodies), and the accounting cluster (continental versus British-American) are significant determinants of certain properties of accounting numbers, such as conservatism and value relevance. As regards value relevance, Hung (2001) provides evidence consistent with the notion that this attribute of accounting earnings is positively associated with shareholder protection.

More importantly, Leuz et al. (2002) find empirical support for their "profit diversion" hypothesis suggesting that the degree of earnings management decreases with the level of investor protection. The reasoning behind this relationship is that strong investor protection effectively restricts insiders' profit diversion (private rent-seeking) and thus reduces their incentive to conceal these activities. In contrast, under weak investor protection, insiders' profit diversion activities are likely to be high, and because there is something to hide, the likelihood of earnings management increases.

Following Leuz et al. (2002), we hypothesize that in countries characterized by weak shareholder protection managers may have more incentives to hide their (earnings decreasing) profit diversion activities through earnings management. More precisely, we hypothesize that when earnings fall just below a critical yardstick (that is a factor of ten) because of profit diversion, managers may have incentive to hide these activities by exercising upward CEM. (5) This managerial behavior may accentuate in settings where the litigation risks associated with private rent-seeking are high. Overall, we thus expect a negative relation between shareholder protection...

NOTE: All illustrations and photos have been removed from this article.



More articles from Journal of International Accounting Research
Some cross-cultural evidence whistle-blowing as an internal control me..., January 01, 2003
Groot, Tom, and Kari Lukka, eds., Cases in Management Accounting: Curr..., January 01, 2003
Sunder, Shyam, and Hidetoshi Yamaji, eds., The Japanese Style of Busin..., January 01, 2003
Epstein, Barry J., and Abbas Ali Mirza, IAS 2002 Interpretation and Ap..., January 01, 2003
Ferdinand A. Gul, Hong Kong Auditing: Economic Theory and Practice.(Bo..., January 01, 2003

Looking for additional articles?
Search our database of over 3 million articles.

Looking for more in-depth information on this industry?
Search our complete database of Industry & Market reports by text, subject, publication name or publication date.

About Goliath
Whether you're looking for sales prospects, competitive information, company analysis or best practices in managing your organization, Goliath can help you meet your business needs.

Our extensive business information databases empower business professionals with both the breadth and depth of credible, authoritative information they need to support their business goals. Whether it be strategic planning, sales prospecting, company research or defining management best practices - Goliath is your leading source for accurate information.